Grid StrAIn: AI & Grid Reliability Part 1—The AI Energy Crunch: How Data Centers Are Reshaping Grid Reliability
Grid StrAIn: AI & Grid Reliability is a new, thought-provoking blog series by Matt Carrara—Industry Leader and Doble Engineering President—exploring how the AI boom is driving new challenges and opportunities for grid reliability.
The first installment, “The AI Energy Crunch: How Data Centers Are Reshaping Grid Reliability” sets the context for how AI-driven energy demand is placing new and unpredictable strains on the power grid. Other blogs in this 5-part series will follow in coming weeks.
Artificial intelligence is fueling a quiet crisis for the power grid. Demand is rising faster than utilities can plan for, and few are ready for what’s coming.
Once confined to labs and startups, artificial intelligence is now embedded in nearly every sector, from manufacturing and finance to defense. But as adoption accelerates, so does its pressure on the energy infrastructure.
Generative AI models and large-scale machine learning systems consume staggering amounts of electricity. The Department of Energy recently reported that data center load growth has tripled over the past decade and is projected to double or triple by 2028. This is not a temporary spike. It’s the beginning of a long-term transformation in how and where energy is consumed. And the grid isn’t built for it.
AI Demand Isn’t Predictable. And That’s a Problem
AI doesn’t behave like traditional industrial load. It doesn’t follow seasonal patterns or fixed cycles. Demand can spike overnight: when a company launches a new AI tool, scales a data center, or upgrades computing infrastructure. These aren’t minor bumps. They’re massive, energy-intensive surges that can hit without warning.
What’s more, these surges aren’t limited to centralized cloud campuses. Google recently introduced the ability to run Gemini AI models directly within their data centers. This means high-density AI workloads are now moving closer to the edge, into facilities that weren’t originally designed to handle such massive computing or energy requirements.
The challenge isn’t just rapid growth. It’s a lack of visibility. Utilities often don’t know when, where, or how much load will appear. Without time to plan, they’re forced into reactive mode–working to stabilize systems that are already stretched thin. In urban areas, where key portions of the grid are more than 50 years old, high-density, high-demand data center builds can create dangerous pockets of instability. And while utilities work to keep pace with AI demand, they’re also navigating a perfect storm of external pressure.
Tariffs, Trade Policy, and Supply Chain Risk Add Fuel to the Fire
As utilities race to prepare for AI’s energy appetite, they’re also facing a storm of external constraints. Tariffs, international trade shifts, and supply chain bottlenecks are slowing access to the transformers, bushings, and substation components utilities need to support grid expansion.
This isn’t just about higher costs. It’s about time. Lead times for critical equipment are increasing, while permitting delays stall infrastructure upgrades. Without proactive investment and greater agility in sourcing and planning, utilities risk falling behind, right when the stakes are highest.
The consequences of this go beyond cost overruns. Delays compounded by unpredictable AI demand and data center expansion increase the risk of brownouts, grid instability, and stalled grid modernization efforts.
The Grid Isn’t Ready, But It Can Be
Solving the AI energy crunch requires more than added capacity. It demands a smarter, faster, more flexible grid that runs on real-time visibility, condition-based diagnostics, and informed decision-making.
But insight alone isn’t enough. As systems modernize, utilities also need the expertise to interpret what the data is telling them. That means investing not only in smarter tools, but in the engineers and field teams who know how to use them. With up to 400,000 U.S. energy sector employees expected to retire within the next decade (McKinsey), the industry must prioritize workforce development alongside infrastructure upgrades. Long-term reliability depends on both innovation and experience: technology and the people behind it.
At Doble, we’ve helped utilities navigate disruptive change for more than a century. And while the technologies have evolved, the mission remains the same: protect the grid, prevent failure, and power a more reliable future.
Additional Information:
- Blogs:
